客獲得数向上のためのタクシー車両配車問題のモデル化と目標エリア決定ルールの構成

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2012/10/01

タイトル(英語): Acquiring Rule-sets for Deciding Target Areas in Distributing Cruising Taxicabs

著者名: 大原　誠（神戸大学），玉置　久（神戸大学）

著者名(英語): Makoto Ohara (Kobe University), Hisashi Tamaki (Kobe University)

キーワード: タクシー配車，遺伝的機械学習，不確実最適化問題　　distributing taxicabs，GBML，uncertain optimization problem

要約(英語): This paper presents a way to acquire effective rule-sets for realizing the suitable distribution of cruising taxicabs based on a decentralized autonomous manner. As a result of the revision of Road Transportation Law, a taxi business have been saturated by new entries, and rookie drivers particularly have a lot of difficulty in finding passengers, because of the lack of the knowledge or experiences. Here, it is desired and effective for less-experienced drivers to design a support system which suggests potential destinations for finding passengers. As for the problem of distributing cruising taxicabs, there are many uncertainties caused by incomplete observations, e.g. locations of passengers as well as other taxicabs. Due to these uncertainties, it is impossible to distribute taxicabs optimally, e.g., by a certain centralized control. This paper proposes a decentralized and autonomous framework for distributing cruising taxicabs by adopting priority-based rules, which calculate the priorities of areas, which are to be adjusted according to the situations of each taxicab. Through some computational simulations, the effectiveness and the potential of the proposed method are confirmed.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.132 No.10 （2012） 特集：医療関連電子技術のニュートピック

本誌掲載ページ: 1645-1655 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/132/10/132_1645/_article/-char/ja/